318 THE PRINCIPLES OF LOGIC BOOK II. PT. II came a paler counterpart. And, when we once discern this fact through the mist of ambiguous and misleading formulae, there is an end to the theory which hides or obscures it.
What was contiguous is now non-existent, and what is " re-instated " has never been contiguous. Let us look at the facts. A sensation A excites by similarity an image b, and, on this, contiguity has to do all the rest. But has b ever been contiguous to anything? In the case before us there are two possibilities. The fact from which we start is this — we have had an impression B along with an impression C, and we have an impression A. Now what are the two possibilities? In the first place it is possible that we never have had a feeble image resembling B. And this is more than possible, for in an early mind it is also probable. But in this case, when A excites an image b, there is absolutely no contiguity of anything with anything. Not one of the supposed elements in our reproduc tion has ever been contiguous with any other; and, this being so, reproduction will not take place. This first possibility ap pears to me to have been overlooked. Let us now pass to the second. We here have had the contiguous impressions B — C. These we suppose to have been followed by one or more pale pairs of images &1 — c1, b2 — c2, &3' — cz. These are all like each other, but they all are realities each of which is not the same as any other. We now experience a sensation A. This also is like the previous sensation we have called B, and is like the images &1, b2, b\ But every one of these, I must beg the reader to remember, is by this time absolutely non-existent. What then is to happen when A is presented? It calls up by similarity an image b4. But this is not what we want. For we want an image b* — c4; and contiguity is invoked to pre sent us with c4. But is invoked in vain. For as yet c* has never existed, and ex hypothesi it is to be made to exist by means of contiguity. On the other hand b* has never been contiguous to anything at all. We have reached once again the old result. There is no association by contiguity. What is called up by association has never been contiguous; and what has been contiguous can not be called up. The contiguity which now operates is a past contiguity, which is not recalled and can not be recalled, but which, according to the pious legend, is somehow passed on like original sin.
CHAP. I THE THEORY OF ASSOCIATION OF IDEAS 319 But if this is so, then Association by Contiguity is exploded finally. No exciting of similars will save it from annihilation. For the similars excited have not been contiguous, and what was really contiguous can not be excited. If present sensations are qualified by images in the way described, still on that (false) hypothesis there is no reproduction by association. There can be no association where the elements are not co existing associates. But if they do already co-exist and thus are associated, then how in the name of all that is miraculous can one bring about the co-existence of the other, and by means of their co-existence?
§ 19. If the school of Experience is in earnest with its principles there can be no such thing as Association. But is it in earnest? Notwithstanding all its public protestations may it not secretly look for the Resurrection of the Body? Does not the charm of Similarity shake the realm of Hades, and conjure from its grave the reluctant past? Is anything too hard for Association? Its spell has prevailed over the mind of its votaries, and, though their lips may deny, yet Associa tion itself has helped their unbelief by its own divine power. They do believe in the miracle of resurrection. But they be lieve blindly and unconsciously, compelled by the strength of a tacit conjunction of meaning with phrase.
We saw that, by the admission of its advanced disciples, association depends upon similarity. If there is no reproduc tion by Similarity, it is admitted that there is no Association at all. I shall now press this consequence. If you do not believe in this kind of Association, you believe in none. But if you do believe in it you believe in a miracle which upsets all law. And furthermore there is no evidence a posteriori to confirm this miracle. In plain words Association by Similarity is a downright fiction. It is not called for by the facts; and it involves besides metaphysical assumptions which I confess stagger me, and which I think may somewhat surprise others. I shall show the reader how the school of Experience has swallowed the most outrageous metaphysical doctrines, and that he must follow their example or leave their company.
§ 20. (ii) Association by similarity, if it is anything at all, is a means of exciting ideas that are not present. If it will not give us what at present, and apart from its agency, we are 32O THE PRINCIPLES OF LOGIC BOOK II. PT. II without, then it surely is a self -condemned fiasco, that is not worth discussing. We may perhaps agree that an agency which recalls and yet recalls not anything but what is already on the spot, is something like a piece of nonsense. And I propose to show that Association by Similarity is this piece of nonsense.
Similarity is a relation. But it is a relation which, strictly speaking, does not exist unless both terms are before the mind. Things may perhaps be the same in certain points although no one sees them; but they can not properly resemble one another, unless they convey the impression of resemblance; and they can not convey it unless they are both before the mind. This is not merely an assertion I have chosen to make.8 Let us see what is told us by J. S. Mill.
" Any objects, whether physical or mental, are related, or are in a relation, to one another, in virtue of any complex state of consciousness into which they both enter" (on James " Likeness and unlikeness are themselves only a matter of feeling: and that when we have two feelings, the feeling of their likeness or unlikeness is inextricably interwoven with the fact of having the feelings. One of the conditions, under which we have feelings, is that they are like and unlike: and in the case of simple feelings, we can not separate the like ness or unlikeness from the feelings themselves. It is by no means certain, however, that when we have two feelings in immediate succession, the feeling.of their likeness is not a third feeling which follows instead of being involved in the " I have two sensations; we will suppose them to be simple ones; two sensations of white, or one sensation of white and another of black. I call the first two sensations like; the last two unlike. What is the fact or phenomenon con stituting the fundamentum of this relation? The two sensa tions first, and then what we call a feeling of resemblance, or of want of resemblance. Let us confine ourselves to the former case. Resemblance is evidently a feeling; a state of the consciousness of the observer" (Logic, I. 75).
Is not this quite plain? Does it leave any doubt? Is it not clear that two mental elements are not like, unless I have CHAP. I THE THEORY Of ASSOCIATION OF IDEAS 32 1 them before me at once or in immediate succession? But, if so, what meaning can we attach to the calling up of an idea by similarity? If the relation does not exist until the idea is called up, how can the idea be called up by the relation? Is it not, the moment we look below the surface, mere verbiage and nonsense?
§21. In the first place what is called up is absolutely non-existent. We are told, not once but again and again, that a feeling gone is gone for ever. And the same thing holds of particular images. If these exist, then the past exists, and the procession in the mind is not real but illusory. Are we to believe this, and believe it in the teeth of our as severations? But if we can not believe it, and if the past does not exist, then we must believe in a relation between the existent and the non-existent; and believe that the whole (relation and relateds) is one state of our minds. If, on the other hand, the past can exist, this miracle will not save us from annihilation. In the relation of similarity both terms must be present, and the fact that one calls up the other by this relation, postulates that one of the terms must be absent. It is therefore both present and absent at once. On either hypothesis we are landed in contradictions; and I have re deemed the promise I gave to the reader. An idea is absent and at the same time present. It is not there and so is brought in by a relation, which relation is nothing if the idea is not there. And a union, which is impossible out of the mind, persists between the existent and what is wholly non existent. Could anything be more insane than this wild meta- physic?
§ 22. But I shall be told " You are deceiving us; it is incredible, it is impossible that our sober countrymen can have been so imposed upon." I answer, That question is easily settled. It is admitted that by " association " they must mean something, and what else do they mean?
The Experience Philosophy has to meet two objections. It has to explain how the non-existent can be related to the existent. And when it has done that, it must explain how the absent can be recalled by the present, when similarity implies common presence and reproduction excludes it. Sup pose that the former difficulty has been slurred over by some 322 THE PRINCIPLES OF LOGIC BOOK II. PT. II metaphysical formula of " the potential and the actual," or some distinction between my mind and other minds, yet the second remains. Suppose that your past series somehow exists, yet how, I ask, are you going to get at it? Mere partial identity of the present and the past would not be what you want, since this would not be an actual relation in your mind.
This is what Maas meant by the following objection. " The mere similarity of two ideas (or sensations) can not possibly be a cause of their association. For similarity is an objective relation of the ideas themselves; while association is a subjective connection in the imagination. But the latter does not follow from the former, nor tend to follow from it " (Versuch, p. 55). By " similarity " Maas of course here meant "partial identity," and his argument is quite simple. The question is, Why does my mind go from one element to another? If you say, it goes because the elements seem like to it — that supposes both to be there. But if you say, it goes because they are like apart from it — then it goes by a miracle, for it is influenced by something which to it is nothing. Sir W. Hamilton (Reid, p. 914) has replied to this argument by a criticism which shows that he did not understand it.
The Experience Philosophy may have a reply to these objections, but I confess I can not anticipate its answer. Per haps it may fall back on a simpler view. It may say, after Wundt (Phys. Psych. 788), "every perception or idea tends to call into consciousness another like itself." As to the truth of this expression I shall have something to say afterwards. But at present I say this. Whatever else it is, it is giving up Association and throwing it overboard. For it is the mere statement of a phenomenon; and it is not an explanation. The entirest belief in the truth of this formula is compatible with the entirest disbelief in the doctrine of Association. We might explain the alleged fact that, given any one element, another like it may come up, by a theory of the spontaneous fission or gemmation of ideas; and this in my opinion would be a theory which, by the side of Association, is sober and rational. We might explain it again by a physiological dis position to a certain cerebral function, which (given the stimulus of a new perception or idea) passes into fact. And CHAP. I THE THEORY OF ASSOCIATION OF IDEAS 323 against this explanation I will not say one word. I will in sist only on this, that it is not a psychological explanation at all, and that in the hands of those who know their own busi ness it is not offered as such. If this is the only possible ex planation, then a psychological explanation is relinquished as impossible, and the Laws of Association as commonly given will not explain anything. Thus the Philosophy of Experi ence must take its choice. It must either rehabilitate its bar barous mythology, or admit that, though the fact of reproduc tion is known, it has no psychological explanation to offer, and is confessedly bankrupt. It has rested its all on reproduction by similarity, and we have shown that this is an impossibility.
§ 23. But our proof no doubt will not cause much disquiet. I shall be told " You can not demonstrate away the facts." And I will therefore proceed to my second contention. The explanation offered is not only impossible, but it is also un called for. There is no evidence for it a posteriori. The facts of reproduction are much better explained on another theory. We have seen this already in our first Book, but I will exhibit it once more.
Let us take a fairly simple instance of reproduction. A young child, or one of the lower animals, is given on Monday a round piece of sugar, eats it and finds it sweet. On Tuesday it sees a square piece of sugar, and proceeds to eat it. In this we have of course volitional phenomena as well as intel lectual, but perhaps we may simplify the case so as to make it serve.
Now on the Association theory how is the fact inter preted? I suppose in some way like this. The presentation to the eye of Tuesday's piece calls up by similarity the idea of Monday's piece. That is a feeble counterpart of the original sensation, and it calls up by contiguity feeble counterparts of Monday's felt movements and Monday's following sweet taste. The fact which ensues is hence the mental presence of Tuesday's perceived square piece, felt to be like another paler imagined round piece, with which latter a whole set of other images come in. Now the conclusion, at which we have to arrive, is the qualification of Tuesday's piece by these images which are attendant on the idea of Monday's piece; and at first sight there seems no way to this 324 THE PRINCIPLES OF LOGIC BOOK II. Pi. II result. For the conclusion is not merely a vicious inference, but it does not even look like a probable mistake. Tuesday's sensation and Monday's image are not only separate facts which, because alike, are therefore not the same; but they differ perceptibly both in quality and environment. What is to lead the mind to take one for the other?
Sudden at this crisis, and in pity at distress, there leaves the heaven with rapid wing a goddess Primitive Credulity.9 Breathing in the ear of the bewildered infant she whispers, The thing which has happened once will happen once more. Sugar was sweet, and sugar will be sweet. And Primitive Credulity is accepted forthwith as the mistress of our life. She leads our steps on the path of experience, until her fal lacies, which can not always be pleasant, at length become suspect. We wake up indignant at the kindly fraud by which the goddess so long has deceived us. So she shakes her wings, and flying to the stars, where there are no philoso phers, leaves us here to the guidance of — I can not think what.
The school has not yet accepted this legend, and I narrate it partly because I am not sure that it is not relevant, but mainly because it has always seemed to me perhaps the most striking of all those creations which we owe to the imagina tion of Professor Bain (Emots. p. 511 and foil.).
§ 24. The less poetical but not less fabulous view would appear to be this. Given a perception A together with an image b, which resembles it and has a train of attendant images c, d, and e — the problem is how to transfer to A the content of e. And what accomplishes the feat is the Law of Obliviscence. This powerful agent obscures everything in the train between A and e; and it also obscures any part of e which is not suitable to A. The residue of e then adheres to A; that is, I think, the two run into one. And so we get the conclusion "This piece of sugar is sweet," by a process which logically may seem rather vicious, but which appears none the less to be the essence of reasoning.* * In the lowest stages of mind this theoretical conclusion of course would not appear. There would be action or attempt without anything like a judgment. The principle however would be exactly the same; and when the theoretical conclusion comes, it must come in this way.
CHAP. I THE THEORY OF ASSOCIATION OF IDEAS 325 I can not say if this statement of the Association doctrine is fair, but I hope it may be so. Let us see what objection we can find to its process.
The main objection is that there is a great deal too much of it. It is much too elaborate for simple phenomena. It first introduces a complication which does not exist; and then, having invented this complication, it removes it by a process which is not real.
It is obviously no fact which we can discover by observa tion, that when Tuesday's sugar is presented to sense, a similar piece or similar pieces come up, in their particularity and with all their differences, before the mind. No one gets such a fact from observation. It is in short a theoretical fiction. I do admit that afterwards, when memory is developing, there is something which can give ground for a mistake of this kind. But then of course reproduction must come before memory, and in the present case we are not concerned with the latter (cf. p. 36). The fact before the mind is that this sugar suggests both sweetness and eating without any images of any other pieces of sugar at all. In the first Book I enlarged on this point by anticipation, and, I confess, it seems to me quite plain.
§ 25. But I shall be told, that although we can not be aware of them, these images exist, and they are removed or adapted by the Laws of Obliviscence. But this process strikes me as another fiction, piled up to support the first fiction against the pressure of experience. I will quote a passage from J. S. Mill.
" The reader...is now...familiar with the...fact,...that when, through the frequent repetition of a series of sensations, the corresponding train of ideas rushes through the mind with extreme rapidity, some of the links are apt to disappear from consciousness as completely as if they had never formed part of the series. It has been a subject of dispute among philosophers which of three things takes place in this case. Do the lost ideas pass through the mind without consciousness? Do they pass consciously through the mind and are they then instantly forgotten? Or do they never come into the mind at all, being, as it were, overleaped and pressed 326 THE PRINCIPLES OF LOGIC BOOK II. PT. II out by the rush of the subsequent ideas?" (on James Mill, The question opened in the above quotation may be stated thus: Given an indirect connection of ideas in the mind, to find the way in which it becomes direct. I do not wish here to enter into this general question. But I must point out that Mr. Mill has raised it in a form which precludes any satis factory solution. For the ideas connected are not really a mere series of particular images, and the fact has thus been perverted beforehand. And if we suppose that, in some ex ceptional case, we have got a mere train of individual images, then not one of the " three things " could possibly be opera tive. For so long as the ideas remained these mere images, no connection at all would be established between them. We may be sure that, whatever in the end may be the detail of the psychological process, one side of it would consist in turn ing these images into universals. And for this reason the Laws of Obliviscence, as we have them stated by Mr. Mill, are fictitious processes. Even if you start with a complica tion and a train of ideas, yet they can not deal with it.
But the point on which I desire to insist, is that in an elementary case of reproduction, such as we are now con sidering, the complication presupposed by these Laws has no existence at all. The data from which they start are pure inventions, and it is hence an impossibility that any one of the suggested " three things " should happen. The fact which Obliviscence postulates is this: A is the sugar calling up by similarity an image of sugar b; and b calls up by contiguity an image of movement c; and c calls up an image d of a particular sweet taste. But this fact does not exist, and the alleged process stands therefore on unreality.
There is in the first place no reason to suppose that this train of ideas, which is presumed to rush through the mind, is a counterpart of the original perception and action. What ground can we have for an assumption that the particular images, b, c, and df are like in all their detail to any train of impressions we ever have had? Admit the train, what reason have you to affirm that there is anything more than a general likeness? What ground have you for the assertion that, if you could look into the past, you would see a train of impres- CHAP. I THE THEORY OF ASSOCIATION OF IDEAS 327 sions B, C, and D, of which these present images are copies? Why must d be an " exact likeness " of the particular pleasant eating D? These dogmas seem to me to be nothing but postulates. The fact, so far as I observe it, shows me that, without respect for the past, such images vary freely within a certain limit, and that this limit is fixed by the universal connection which appears in all of them. But, if so, then what is associated is not particular images. The universal which has been deposited is the active principle, and the par ticular images as such are quite inert.
And in the second place the alleged process imports an other gross fiction into the data. It tells us that similarity calls up an image b, which is a copy of Monday's piece of sugar. We have just seen that, if present, the image need be no copy: and now we go further. For in our elementary case the image b has no existence. I repeat once more that it is a pure invention, necessary for the theory but absent from the fact. When Tuesday's piece of sugar is present, the attributes of whiteness and crystalline appearance reproduce the ideas of movement and sweet taste, without any such link as another and different piece of sugar. It is not merely that we can not find such an image now. We never could have found it. It never has been there. And we need not ask at length if the Laws of Obliviscence could serve to obscure it, unless some evidence is produced to show that it is more than a mere chimaera.
And, as we have seen, it is a chimsera that will not work. For when you have got your image of Monday's sugar, you are left precisely where you were before. You have got an element which has just been born, and which therefore can never have been contiguous to anything in its life. And if you say " But it resembles what was contiguous; " then this is not only to desert your principles, but it also tends to expose you to ridicule. If you want what is the former piece of sugar, you can not get it. But if you want what is like the former piece, then you have it already in the present per ception.
Your fictions do not help you, and why should you cherish them? Why invent the existence of similar images which lure the unwary to vicious inferences? Why suppose that 328 THE PRINCIPLES OF LOGIC BOOK II. PT. II " trains of ideas," of which the mind knows nothing, float across it in procession, and then go on to manufacture a Law of Obliviscence which ties a bandage over its eyes? Because, if you do not, you are forced to admit that the mind does not go always from particulars to particulars, that indeed it never can go from particulars direct to particulars, that in short the Experience psychology is exploded.
§ 26. Let us give once more the natural interpretation of the simple fact. The natural view is that Monday's experience remains in the mind, not in the shape of particular images, but as a connection between elements of content. This is a result which in its metaphysical nature we can not here char acterize,10 but, in its appearance to us, it is easy to describe. It is a tendency to pass from one universal to another, when ever the first of these is presented in an actual perception or image. In the instance we are examining, the shape, the size, the person giving, the where, the when, and the how have all gone. Nothing is left but a tendency to pass from element to element, from whiteness and crystalline appearance and hard ness to eating and sweetness.
Monday's experience, let us say, has established the con nection " white-eaten-sweet." On Tuesday " white " is given, and so we have " this-white." We advance by means of an elementary synthesis to " this-white-eaten-sweet," and, ignor ing that part which does not interest us, we get " this-eaten- sweet," or, elliptically, " this-sweet." I grant you the " sweet " is now fully particular, but its particularity has had nothing to do with its recall. On the contrary its detail depends upon the context which has recalled it. And there is no particular image of " white " at all; for the universal " white " is what has worked, and that of course was given in the present perception.
Where is Similarity here? It does not exist. Similarity implies the feeling of diversity, and here the difference of particulars never comes before the mind; it is in no sense present.
Let us give up Similarity -£ Contiguity -(- Obliviscence or Primitive Credulity. Let us postulate Identity + Contiguity, and then all is easy. But there are two things we must re member. The contiguity is a connection of universals, and is CHAP. I THE THEORY OF ASSOCIATION OF IDEAS 329 therefore not the contiguity of the Association school. And the identity is not present to the mind. The mind, if you keep to simple cases, knows nothing of any difference. It goes straight from what is given to an additional fact.
§ 27. Let us state our view as a working hypothesis, something that need not be true or even possible. Let it be granted there is a mind X with certain functions; let it be granted that X may be stimulated to perform again any function which it ever has performed; let it be granted that in every function there is a connection of elements, as a-b; let it be granted that presence of a tends to excite X to per form again the function which contains a-b; then let a be given in a fresh context, as Ca. On this X is stimulated to go on to b thus, Car-b; and the product Cab now comes before the mind — which is the fact to be explained. If this ex planation is false, admit at least that it is simple.
We are asked to believe it is more in accordance with " experience " to say, Similarity is a tertium quid ensuing only on the presence of a pair of elements, and, when but one is present, Similarity brings the other. It is " science " when we asseverate that mental phenomena are realities which can exist only while they are perceived, and then speak of " recalling " them, as if they were ambassadors on foreign employment, or " calling them up " as though they were servants in the kitchen, and as if " relations " were wires that rang the bell, or were fishing-lines baited with similarity to draw up from non- existence the ghosts of the past. It is " positive knowledge " to make that come before the mind which does not come be fore the mind, and then to remove it by a fictitious expedient. Yes, sooner than run the risk of believing in metaphysics, there is no superstition so gross, no mythology so preposterous that we ought not to believe in it, and believe anything sooner than cease to believe in it.
§28. But what is it that forces us to these desperate shifts? Not the facts themselves, for we violate them. It is simply the shrinking, as we think, from metaphysics. And this, after all, is nothing but metaphysics. It is our unreason- ing fidelity to a metaphysical dogma which has driven us to adopt these embarrassing results. For why is it we are so sure that identity is impossible, and that a synthesis of urn- 33O THE PRINCIPLES OF LOGIC BOOK II. PT. II versals is a " survival " of superstitions, which in the nine teenth century are out of date? It is because we are sure that there can be no reality but particular existences, and no mental connection but a relation of these units; and that hence identity is not possible. But this is of course a meta physical view, and, what is more, it is nothing but a dogma. The Philosophers of Experience have, so far as I know, never offered any proof of it; they have heard it from their fathers, and their fathers had heard it. It is held true because of the continuity of tradition in a Church, which must have truth, since it has never failed to preserve its continuity. Has the school ever tried to support it by any mere rational con siderations?